.WAV audio files

Cha-Ching!!
Post Reply
musiconradio.com
Posts: 146
Joined: Thu Mar 29, 2012 12:24 pm
Location: Livingston Alabama
Contact:

.WAV audio files

Post by musiconradio.com » Mon May 19, 2014 6:10 pm

Hi everyone,

In a word. The .wav audio files form the record companies are LOUD. Compared to normal cuts. (most were ripped from CD's) We use Audition 1.5.

Would you:

Normalize the files (both channels independent)

Or use amplification to bring down the level.

or any other suggestions.

Looking to keep the quality.

Thanks.

User avatar
davek
Posts: 166
Joined: Sun Aug 02, 2009 7:28 pm
Location: Sydney, Australia

Re: .WAV audio files

Post by davek » Tue May 20, 2014 1:21 am

What you want to do is run an analysis to find the average or RMS level of the track. You will get a separate average level for L and R, so sum those number and divide by 2 to get the average mono level.

Say for example the average level is -3.4dB. You want to aim to bring that average level down to where you want it. If you are aiming for -10dB then attenuate BOTH channels by 6.6dB using the amplification routine.

The problem with just using normalize is it will pick the highest peak and use that as the maximum level, which might bring your average level down too much...
Last edited by davek on Thu Sep 11, 2014 4:09 pm, edited 1 time in total.

User avatar
Dale H. Cook
Posts: 773
Joined: Thu Dec 20, 2007 9:08 am
Location: Roanoke/Lynchburg, VA
Contact:

Re: .WAV audio files

Post by Dale H. Cook » Wed May 21, 2014 6:05 am

musiconradio.com wrote:In a word. The .wav audio files form the record companies are LOUD. Compared to normal cuts. (most were ripped from CD's)
Much current popular music is hyperprocessed, and the ripped cuts you are comparing them to are probably not hyperprocessed. In Audition, do the recent cuts that are too loud look like a freshly-mown lawn, with all of the peaks at about -1 dBfs? If so, they are undoubtedly hyperprocessed. My approach is to use Bob Orban's loudness meter software to measure some of the loudest sounding (but acceptable) library cuts to establish a baseline, and then normalize the hyperprocessed cuts to match that reference loudness. If your processing is not overly aggressive that will help, but processing in general still allow the hyperprocessed cuts to sound louder on the air than older cuts. I have tried a few software expansion systems to try to undo some of the hyperprocessing with mixed results. I have one recent album by an artist whose earliest albums had wonderful dynamics, and I cannot listen to because its lack of dynamics is very tiring to my brain. I have found "remastered" CDs of older material that were also hyperprocessed that I also cannot listen to because the tracks don't have the original dynamics.
Dale H. Cook, Contract Engineer, Roanoke/Lynchburg, VA
http://plymouthcolony.net/starcityeng/index.html

User avatar
Shane
Posts: 771
Joined: Fri Feb 01, 2008 12:08 am
Location: Omaha
Contact:

Re: .WAV audio files

Post by Shane » Sat May 24, 2014 11:40 am

Can Audition 1.5 do such an analysis? Another question, the answer to which escapes me, is: why does it come out differently (without the lawn mower effect) if you take the same material and "dub" it by recording it in real time?

In answer to one of the OP's questions: there isn't any difference between normalizing and amplification except that normalizing takes the loudest bits and sets the file level of everything else based on that, with no change in the dynamic range unless you use a very low percentage or dB level that puts the softest parts in the mud. Amplify does the same thing except you have to know (or guess repeatedly using undo) the correct level you want.

Here's another way. Play back your ripped cut into something with a real VU meter on it and see where the average level is. (The meters on the Logitek series of consoles are good for this because they read up to +20.) Say, the average is +10 on the meter. Now you can just use amplify to take the entire cut down 10dB. This way you are using the average instead of the instantaneous peaks to set the overall level.

However, when using this method where you find you need to ADD gain, you will need to check that no instantaneous peaks go over 0dBfs on the waveform screen. A recording with lots of dynamic range (I call them "peaky") will be the most likely to cause this. In that case, in order to make the cut more uniform in average level, you might consider adding a little compression first or at minimum a brick wall limit to take just the soaring peaks down. With light compression you can preserve some semblance of dynamic range. With limiting only the very loudest instances, you preserve all of the original dynamic range except at those places where the level rises to the brick wall.

To accomplish that you would have to do the amplification as determined by the VU meter first, see if any peaks go past 100%, then undo the amplification, do whatever dynamics processing you think is needed, and then reapply your determined amplification. If you don't undo first, you will have a distorted peak that will still be distorted after you limit it.
Mike Shane, CBRE
---Omaha---

User avatar
Dale H. Cook
Posts: 773
Joined: Thu Dec 20, 2007 9:08 am
Location: Roanoke/Lynchburg, VA
Contact:

Re: .WAV audio files

Post by Dale H. Cook » Sat May 24, 2014 12:37 pm

Shane wrote:Can Audition 1.5 do such an analysis?
No - it does not incorporate loudness measurement. Loudness is a subjective quantity, and over the decades there have been many electronic systems for measuring it. The earliest of which I am aware is the system implemented in the CBS loudness monitor, introduced by CBS Laboratories in a paper presented at the 1967 NAB engineering conference. The Orban loudness meter program incorporates a version of the CBS system, as well as the ITU BS.1770 standard.
Shane wrote:why does it come out differently (without the lawn mower effect) if you take the same material and "dub" it by recording it in real time?
I am not quite sure what you mean. If you take a hyperprocessed WAV file, play it back through a professional audio card, re-record it as a WAV through a professional audio card, and normalize the resulting WAV to the same value as the original WAV, the old and new WAV files should be very close to each other.
Shane wrote:... into something with a real VU meter on it and see where the average level is.
That, however, will probably not correspond to measurements with a loudness meter when dealing with material that has had its loudness significantly increased with the use of analog or digital processing.
Dale H. Cook, Contract Engineer, Roanoke/Lynchburg, VA
http://plymouthcolony.net/starcityeng/index.html

User avatar
Deep Thought
Posts: 3217
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: .WAV audio files

Post by Deep Thought » Sat May 24, 2014 1:03 pm

Unfortunately, you can't "unprocess" severely-processed audio no matter what the software or hardware vendor claims. Once it is squashed like that the original source material is unrecoverable. You can make it less offensive with a LOT of experimenting but that usually just results in a different set of distortion problems, and the time to get it to that point is not cost-effective.

Getting back to the original question, see if you can compare the peak-to-average ratio of the cuts you think are too loud vs. the ones you ripped from CD which you think are OK. If the ratio is similar then attenuation of the file levels will help, but I would suspect that the loud ones are severely clipped at the 0 dBfs digital maximum level point, which is unrecoverable.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

TPT
Posts: 682
Joined: Mon Dec 03, 2007 3:18 pm
Location: St. Marys, WV

Re: .WAV audio files

Post by TPT » Sat May 24, 2014 1:27 pm

Real-time playback through a good sound card- then-through a equalizer- could be used to make the cut more presentable. Granted, can't un-break an egg, or un-clip smashed audio, but you can re-record at a lower over-all level so that cut doesn't overwhelm the rest of the music library on playback. The EQ can be used to make the original less "mid-rangy" or defuse those too explosive synthesized low notes that can bug your audio processor. In the end it's damage control.

Post Reply